Laser Micrometer Uncertainty - Calibrating Deltronic Pin Gages

[Charles Wathen]

Hi guys,
I have a question on a uncertainty calculation that I'm trying to determine.

I want to determine the uncertainty of a measurement using my Zygo Gold laser micrometer when we calibrate Deltronic Pin Gages. Here is the info:
Zygo laser micrometer resolution: .000001"
Zygo laser micrometer accuracy: .00002"
Zygo laser micrometer repeatability: .000005"
Pin Gage Standard XXX (used to cal laser mic): .00001"

By calculating the above, I get an expanded uncertainty of .000026".

My question is: should I be including the accuracy of the Deltronic Pin Gage I'm calibrating? It has a standard accuracy of +.00004 -0".

If I do, then my expanded uncertainty is .000053" which seems higher than I expected. If I do use this, should I subtract this value from the .00004" tolerance of the Deltronic Pin Gage to come up with an uncertainty of .000013"?

Please help a confused soul.

[Al Dyer]

My advice would be to drop the pin from the laser mike uncertainty calculation.

The laser is being reviewed, not the pin, and you should be able to use any size pin and get the same uncertainty.

Try a test with increasing sized pins and see if the linearity of the laser is stable

[Graeme]

Charles Wathen said:
My question is: should I be including the accuracy of the Deltronic Pin Gage I'm calibrating?

The measurement uncertainty applies to the measuring system. A measuring system is independent of the items it measures.

The uncertainty is used to help determine if the object being calibrated (the pin gage) is within its own accuracy specification.

The ratio between the measurement uncertainty (of the measuring system) and the accuracy specification (of the object being calibrated) is an indicator of the ability of the system to make that particular measurement.

If the object being calibrated was included as part of the measurement system, then it would be reasonable to be required to perform an MSA for every workload item! (Impractical, outrageous and uneconomic, but weirdly reasonable. )

[Ryan Wilde]

Charles Wathen said:

Hi guys,
I have a question on a uncertainty calculation that I'm trying to determine.

I want to determine the uncertainty of a measurement using my Zygo Gold laser micrometer when we calibrate Deltronic Pin Gages. Here is the info:
Zygo laser micrometer resolution: .000001"
Zygo laser micrometer accuracy: .00002"
Zygo laser micrometer repeatability: .000005"
Pin Gage Standard XXX (used to cal laser mic): .00001"

By calculating the above, I get an expanded uncertainty of .000026".

Using all rectangular distributions, I got about the same. (0.0000265")

My question is: should I be including the accuracy of the Deltronic Pin Gage I'm calibrating? It has a standard accuracy of +.00004 -0".

You wouldn't take into account the accuracy of the Deltronic Pin Gage, especially since that is what you are trying to determine. You would, however, need to incorporate the repeatability of measuring a deltronic gauge. Using your system, I would guess this to be around 0.000005" with a single standard deviation.

At this point, we have an uncertainty of about 0.000028".

What we are forgetting is the environmental effects. Temperature is the big wildcard on all dimensional measurements. If your laboratory is 68 ±5 °F (20 ± 2.8 °C), and you don't compensate for the thermal effects, you will have a very different uncertainty.

Thermal expansion of tool steel (such as a Deltronic Pin) is 6.4 µin/in/°F (11.5 µin/in/°C).

Therefore, if you are measuring a 0.5" pin at 72 °F (22.2 °C), and the pin is exactly 0.5" at standard temperature (68 °F/ 20 °C), you would actually measure [0.5" + (0.0000064" *0.5" *4 °F)], or 0.500013". With everything else in the system being perfect (which it isn't), you have an error of 0.000013" due to temperature alone.

I do not know the temperature effect on the Zygo laser, but thermal expansion of the pin, and thermal effects on the Zygo laser must be added to the uncertainty budget.

If I do, then my expanded uncertainty is .000053" which seems higher than I expected. If I do use this, should I subtract this value from the .00004" tolerance of the Deltronic Pin Gage to come up with an uncertainty of .000013"?

No, the tolerance of the Deltronic pin has no bearing on your measurement. The tolerance of the pin under test is entirely arbitrary to the uncertainty of measuring the pin.

Ryan

[Charles Wathen]

Thanks everyone for your feedback.

[Marc]

Also see this thread for some thoughts on measurement uncertainty. It's a relative oldie (check some of the post dates), but it's a goodie:

Types of Measurement Uncertainty of Different Devices - Is there any Standard?